Japanese readings of Sinographic names

« previous post | next post »

Eoin Cullen wrote in:

I recently learnt that although Taipei  たいぺい is generally used as the Japanese reading for Taipei 台北, NHK still uses the colonial form Taihoku たいほく.  Is this still true in 2018? Why would the national broadcaster persist in using an archaic term? To me, it seems it would be comparable to the BBC insisting on using the name Ceylon to refer to Sri Lanka.

I asked several colleagues who are specialists on Japanese what the significance of this usage might be.

Frank Chance:

Most Chinese city names, as most Chinese [VHM:  Sinographic] names in general, are pronounced using the Japanese pronunciations of the characters for the name.   Mao Zedong thus becomes Mō Takutō, for example, and Suzhou becomes Soshū, Kyŏngju is Keishū, Taibei is Taihoku.  A few of the best known Chinese cities get semi-Chinese pronunciations, such as Shanghai becoming Shanhai, Nanjing becoming Nankin, and yes, Beijing becomes Pekin. Hong Kong loses the final g’s to become Honkon, but Guangzhou (Canton) becomes Kōshū,  I don't think it has to do with colonialism but with pronouncing the characters as they are pronounced in standard Japanese in all but the most commonly used names.

Cecilia Segawa Seigle:

I myself don't know what is correct, and the Japanese don't know either. NHK answers they read it Taihoku because it is the historical Japanese reading.

There is no logical explanation and there is no consistency.  I sometimes call Taipei タイホク [Taihoku] and some other times, タイペイ [Taipei].  I am not consistent, because there is no strict rule. If the government rules that we have to be consistent, then we will stick to one reading.

Sorry for being so sloppy, but that's the reality.

Hiroko Sherry:

The funny thing is that when we say Taipei, it is usually written in Katakana タイペイ.

Nathan Hopson:

[VHM:  This section is for readers who are already familiar with the Japanese writing system, so I am not providing Romanizations of kanji and kana.]

The problem is not just with Taipei, though it's certainly a prominent example. Almost all Chinese (in the broadest sense) proper nouns (personal names, toponyms, etc.) are just read in Japanese. If anything, Taipei is an exception to that rule, along with Shanghai (シャンハイ) and Beijing (ペキン). So:

基隆:きりゅう or キールン
高雄:たかお or カオシュン
桃園:ももぞの or… etc.

If I understand correctly, some of this probably has to do with the way that toponyms were rewritten under Japanese rule and then again after 1945:

打狗 (ターカオ) → 高雄 (たかお) → 高雄 (Kaohsiung)
鶏籠 (チーロン) → 基隆 (きりゅう) → 基隆 (Keelung)

I think current pronunciation is a mix of political and pragmatic. For most, it's just easier. For a few (how many, I don't know), it's probably a colonial legacy.

The situation is different with Korea:

Kim Jong-un 金正恩 is キム・ジョンウン (not きんしょうおん), though even as recently as his father, I feel like I saw both きん‐しょうにち and キム=ジョンイル as glosses.

Seoul is definitely ソウル (not 京城 けいじょう anymore), like 釜山 is definitely ぶさん, not かまやま.

How many of us still say Bombay instead of Mumbai, Burma instead of Myanmar, and Cambodia instead of Kampuchea?  I must confess that I still say Peking instead of Beijing from time to time — especially when I'm talking a kind of savory duck, a cute type of dog, or a very special Sinitic topolect.

[Thanks to Tomoko Takami]


  1. Guan Yang said,

    September 26, 2018 @ 9:29 am

    Some institutions still have “Peking” in their official English-language names, such as Peking University and Peking Union Medical College.

  2. Jim said,

    September 26, 2018 @ 9:34 am

    Similarly, Chinese generally pronounce Japanese names according to the Han Chinese pronunciation of the characters, for example Tokyo (东京)as "doong jeeng", Osaka (大阪) as Dah Bahn.

  3. J.W. Brewer said,

    September 26, 2018 @ 12:27 pm

    As to the island on which Taihoku is located, I guess it's just dumb luck that "taiwan" is apparently a plausible Japanese pronunciation of the characters pronounced "taiwan" in Mandarin? (Although maybe "daiwan" would be slightly more plausible in Japanese?)

  4. Ken said,

    September 26, 2018 @ 2:12 pm

    Are we sure that non-standard, ad hoc Japanese pronunciations like Taipei, Honkon, Nankin, and Pekin are not based on the English pronunciations derived from the old postal atlas spellings? Why else would the syllable bei 北 end up as pe in Pekin but as pei in Taipei, except for the fact that they are Peking and Taipei in English? Why else would the syllable jing 京 in Pekin end up as kin in Japanese instead of jin or chin? Is Honkon really the closest you can get in Japanese to the Cantonese pronunciation Heunggong, or is it just the closest you can get to the English name Hong Kong? Is Amoi (the Japanese name for Xiamen, according to Wikipedia) the closest you can get in Japanese to the Hokkienese pronunciation E-mng, or is it based on the English name Amoy? I suspect that this is not Japanese being brought closer to Chinese pronunciations, it is Japanese being brought closer to English ones.

  5. Michael Watts said,

    September 26, 2018 @ 2:22 pm

    Hong Kong loses the final g’s to become Honkon

    Wouldn't the ン assimilate to the following /k/ to produce the same [ŋ] sound that was supposedly "lost" from Hong?

    the articulation points of 'j' and 'k' are very close

    …in what sense? One is palatal and one is velar.

  6. Ken said,

    September 26, 2018 @ 2:25 pm

    Japanese has some funkiness in the way it treats historical Chinese names too. The last two dynasties are Min (Ming 明) and Shin (Qing 清) instead of Mei and Sei, and if you study the Sui and Tang period there are a bunch of names that get plucked out for special treatment (Kutoojo for 旧唐书 and Zui Youdai for 隋炀帝, for example) based (if memory serves) on Go readings of the characters, even while other historical contemporaries get the more usual pronunciations.

  7. Scott P. said,

    September 26, 2018 @ 2:53 pm

    Saying Bombay instead of Mumbai is like saying Munich instead of München, or Seville instead of Sevilla.

  8. Anonymous Coward said,

    September 26, 2018 @ 3:14 pm

    Keelung is always きいるん, even during the colonial times. No idea why.

  9. Anonymous Coward said,

    September 26, 2018 @ 3:27 pm

    釜山 was never かまやま either. The Japanese reading was ふざん, one dakuten away from the current reading ぶさん.

  10. Bob Ladd said,

    September 26, 2018 @ 3:47 pm

    @Scott P. Language Log featured quite a discussion of the whole Mumbai/Bombay question and the Munich/München analogy some years ago. My guess is that not everyone will agree with your view here…

  11. Philip Taylor said,

    September 26, 2018 @ 4:50 pm

    Nathan ("How many of us still say Bombay instead of Mumbai, Burma instead of Myanmar, and Cambodia instead of Kampuchea?") I confess that I certainly use the first one all the time (as also Calcutta, and whatever the other Indian place is that has changed its name completely), I would seek out "[Chinthe] Burmese lime pickle" and "Pekin[g] duck" but visit Myanmar & Beijing, and when I used Kampuchea in writing (following my wife's usage — she is Vietnamese) and then checked, I discovered that the correct name is in fact the Kingdom of Cambodia. But I didn't bother to correct myself. And I do try to pronounce Sri Lanka correctly …

  12. Jim Breen said,

    September 26, 2018 @ 5:33 pm

    Getting back to 台北, I just checked several dictionaries. Most have entries for both readings, with the J-Es usually having タイペイ as the main entry and たいほく as a cross-reference. The J-Js do it the other way round.

    In Jmdict we only had タイペイ until last year, when たいほく was added.

  13. Jerry Friedman said,

    September 26, 2018 @ 6:30 pm

    Philip Taylor: Chennai, formerly Madras, is the one that changed the most.

  14. Toby said,

    September 26, 2018 @ 7:16 pm

    "Mumbai" has too many Hindutva connotations for my liking.

  15. AG said,

    September 27, 2018 @ 1:07 am

    I checked a few northern places like Mukden and Dalian on JP wikipedia, and they had the "colonial" name in hiragana, followed by the modern Chinese pronunciation in katakana. Then I checked a bunch of Korean places, & as described above they seemed mostly in katakana only. Not sure what any of that indicates.

  16. John Rohsenow said,

    September 27, 2018 @ 2:02 am

    This discussion of the pronunciation of place names is just one case of names and terms borrowed via sinographs between China and Japan (and South Korea). I first encountered this problem when compiling the bibliography for my Chinese English Dictionary of Enigmatic Folk Similes (Xiehouyu) when I purchased a Tokyo bookstore a Japanese book on the subject entitled: 歇後語匯編 (sic; 'A Collection of Enigmatic Folk Similes'). My problem came in trying figure out how to pronounce the title of this book in Japanese, and — more importantly — now to romanize this Japanese title in my English bibliograpy. Neither the book store owner nor any other Japanese I asked could help me. 歇後語 is a rather specialized Chinese term (for which I -for better or worse – coined the term 'enigmatic folk similes" (I think Victor prefers 'truncated witticisms')and the spoken term 'xiehouyu' is (or at least at that time was) not generally understood by most non-Chinese sinologists nor in fact by many native Mandarin speakers hearing (or even reading ) it, so it was even more difficult for Japanese to tell me how to pronounce those Chinese characters in Japanese. (I finally settled on Ketsugogo I Hen, although I'm still not sure that's "correct'.)But the general question remains, when terms are borrowed as Chinese based characters/hanzi/kanji, there has to be some agreement as to how they are going to be read aloud in the host language, and that –as several commentators have noted — ends up being negotiated as a function of a number of social-historical-political factors.
    Lastly I am once again reminded of the interesting (?) fact that we seem to have imported a number of Japanese terms "wholesale", "as is" from Japanese (e.g. Noh, kabuki, origami, ikebana, kimono) into English, but we still refer to Chinese operatic type performances as "Peking/Beijing opera"-inaccurately in the opinion of Prof. Elizabeth Wichmann at the Univ of Hawai'i, who would prefer that we simply say 'jingju" in English, just as we say Kabuki and Noh. But of course unless she can start a widespread revolution, the Mandarin term "jingju" is not going to be generally recognized or understood.

  17. Vilinthril said,

    September 27, 2018 @ 2:20 am

    “Cambodia” is not really a relevant example here, as it's the official English name. I'd argue that Timor-Leste, Cabo Verde and Côte d'Ivoire would be better analogies.

  18. DMT said,

    September 27, 2018 @ 5:43 am

    I have asked various Japanese friends about the contrast between the names of contemporary Chinese figures, which usually get pronounced as on-yomi, and contemporary Korean figures, which usually get pronounced as katakana-Korean. Usually people just shrug; the one explanation I have been offered is that some Koreans are offended by the use of onyomi readings, while Chinese speakers don't really care either way.

  19. Rodger C said,

    September 27, 2018 @ 7:06 am

    Well, in Korea onyomi readings are colonial, whereas in China they're simply how those folks talk.

  20. DocRock said,

    September 27, 2018 @ 7:29 am

    When I was you and in Japan in the early 1960's, almost invariably the place names would be read according to the Sino-Japanese readings. Later on, especially in mass media, it has become increasingly normal to see, rather, these other Katakana or Romaji readings. But, after all, we still call Wien "Vienna"; Roma, "Rome"; and Koeln, "Cologne;" etc.

  21. DocRock said,

    September 27, 2018 @ 7:31 am

    "young" vice "you"!

  22. derek said,

    September 27, 2018 @ 8:31 am

    Many French, or French-origin, city names are pronounced using the English pronunciations of the characters for the name. Pa-riss, Or-leenz, Dee-troight…

    The French name for London, "Londres", is a different animal rather than a "pronunciation of the characters" issue.

  23. ajay said,

    September 27, 2018 @ 10:12 am

    Chennai, formerly Madras, is the one that changed the most.

    Some remain resolutely colonial; McCluskieganj comes to mind. Abbottabad and Cox's Bazar are in Pakistan and Bangladesh respectively. Sekunderabad and Kandahar hark back to an earlier period of Western rule.

  24. Philip Taylor said,

    September 28, 2018 @ 3:57 am

    DocRock ("We still call Wien "Vienna"; Roma, "Rome"; and Köln, "Cologne;" etc."). Not all of us.

  25. Guan Yang said,

    September 28, 2018 @ 9:36 am

    Seoul 서울 is an interesting case because unlike most Korean place names, it didn’t have a name in Chinese characters. So for decades, the city was called 汉城 (Hànchéng) in Chinese, its former name, until the government decided it should be the phonetic 首尔 (Shǒu'ěr), which is what most Chinese people now call it.

  26. Philip Taylor said,

    September 29, 2018 @ 5:47 pm

    Jerry Friedman ("Chennai") — Thank you. But I have yet to be offered a "Chicken Chennai" in a (British) Indian restaurant, or to find "Chennai curry powder" on the shelves of a (British) Indian grocer !

RSS feed for comments on this post